Is there a crack in the fundamental assumptions of the AI Safety field? The talk "The Devil You Know" reflects on the limitations of human understanding in predicting the behavior of advanced general intelligence (AGI). I discuss Nick Bostrom's work, particularly his warning against anthropomorphism in AI and his theses about AI behavior. Bostrom's Instrumental Convergence Thesis is critiqued for projecting human rationality onto AI. I argue that current AI research often reflects the biases and philosophies of its human creators rather than the true nature of AI.

Ruby Thelot is a designer, cyberethnographer and artist based in New York. He is an adjunct professor of Design, Technology and Media Theory at NYU. He is the founder of the award-winning creative research and design studio 13101401 inc. His work focuses on the interactions between humans and artificial intelligence, the metaverse and the implications of being-on-line. He has given talks and shown works in Tallinn, Berlin and Abuja, amongst other places. He holds a Masters in Design and Technology from the Parsons School of Design and a Bachelor in Business from McGill University.

AI Ignite is a series of events exploring the field of AI. This event was sponsored by Microsoft and Tola Capital.

Ignite Talks is a global speaker series and talk format. Every Ignite Talk is exactly five minutes long, with twenty slides that auto-advance every fifteen seconds. The unique constraints require an increased level of creativity from the speaker and add an element of excitement for the audience. Ignite Talks events have been held in over 350 cities on six continents and have sparked tens of thousands of talks.

If you liked the video, please subscribe to our channel here: [ Ссылка ]

Follow Ignite on Twitter: [ Ссылка ]
Follow Ignite on Facebook: [ Ссылка ]

свернуть